The first Dungeon Assault utilises a key game mechanic that we want to build on here at FunOrb: the idea of an active offence and a reactive defence. Specifically, your raiders are the active, offensive element and your dungeon is the reactive, defensive element. This key principle of the game works well and was a real departure for us in terms of the technology required to make it work. Dungeon Assault was the first FunOrb game to keep a record of players' profiles active when they were offline.
Dungeon Assault was a very welcome addition to FunOrb and has provided a lot of gameplay hours for a lot of players. It’s great fun. Personally, though, I am also a little bit disappointed with it. In hindsight, I know it is not all we could have delivered. So, back in January, we decided to give it what it deserves in the form of a sequel!
Our first objective was to sit down and discuss what the original doesn't do so well. "It's too random" was a very common statement, as was "it's not deep enough". We also knew that the game needed to be more engaging and involve more strategy. Finally, Dungeon Assault had a board game feel and, while this worked to a degree, it wasn't quite the effect we'd hoped for.
We then moved on to what the original game does well. We got back statements like "it's quick to play", "there's lots to unlock", "stuff happens when you're not there" and "it can be very suspenseful". Thus we had a plan for our prototype: to make it deep with more strategy and offer real interaction, but keep it quick to play, with lots to unlock, a suspenseful feel and things happening when you're not there.
We've now planned out the development path that the sequel needs to have and formulated the design documentation that the developer will work to. We've prototyped a lot of game mechanics and display elements and come up with a more engaging, less random and far more in-depth game than the original.
We've also worked to break down the problem that Dungeon Assault has when the 'end game' is reached: primarily the fact that it doesn't seem to have one. It’s taken a few months to formulate the core design of the game and build the working prototypes that the developer is now building upon, and we feel that it is sufficiently solid in its design for us to start talking about it.
In the following diaries I'll hand you over to Mod Vile, the developer in charge of making this vision a reality. He'll be talking to you about how we plan to change the dungeon creation system, the raid mechanic, how you will be able to control the actions of raiders individually, and generally about how we are trying to make Dungeon Assault II everything that the original concept deserves to be.
The Dungeon Assault II project began with me and Mod Korpz prototyping a new dungeon building mechanic. We felt that management in the original game was too limited and we'd like to give players a chance to lay down their very own dungeon floors. You will now have more say in how a raiding party gets from your dungeon's entrance to its hoard room.
This awesome new feature comes with a price: we had to make sure that we could devise ways to prevent people building boring, predictable dungeons. Not only that, we had to make sure that it would be hard to build dungeons that would be easy to repeatedly raid.
We began with a couple of sessions around a whiteboard to try to establish some concepts for how this new mechanic was going to work. The four dungeon entrances of yesteryear are now gone - a dungeon now has a single entrance located in the centre of the map. To try to make raids more varied, we decided that the player would have reasonable freedom in placing their hoard room.
Next we began to experiment with pen and paper prototypes to get a feel for how this new building mechanic was going to work. We even cannibalised a board game we have to see if we could build an interesting layout. Each new dungeon floor starts with its entrance crossroad tile located in the centre. The player is given one tile at a time to place in their dungeon. The tiles will occasionally have traps or monsters already placed on them from a pre-allocated budget. This is to ensure that everyone's dungeon starts out with roughly the same difficulty level and provides a challenge.
We then had to work out a set of rules that would decide how new tiles would interact with the ones already set down. It was very important to guarantee that it was impossible to build dungeons with unreachable parts. We had to make sure that every new tile could be reached from any existing tile (by knocking through walls if necessary) while at the same time trying to make sure the dungeon didn't become too connected (crossroads everywhere). The answer turned out to be to leave existing doors and walls as much as possible.
To prevent people building dungeons that are spiral in shape and boring to raid, we decided to adopt a 'concentric rings' approach to the construction. When you start building a dungeon you may only place tiles in the eight squares neighbouring the centre tile. Once these squares have been filled in you may then move outward and place tiles in the ring of squares surrounding the nine centre tiles. This process continues until the dungeon is filled. The proportion of tile shapes you receive is also an important factor; too many crossroad or t-junction tiles and the dungeon will become too connected. Where you can place the hoard room is also subject to some restrictions. It shouldn't be easily reachable from the dungeon entrance; dungeons are meant to provide some sort of challenge after all!
After a few experiments, the prototype was coded up. It was important to find out whether the pen and paper model would translate well into a videogame. It quickly became apparent that while the building game was indeed fun, there were a few features that could improve playability. The builder had to show the player how each new tile was going to affect the existing portion of the dungeon, otherwise the rules explained above appeared too random and even unfair. It also became obvious that the new 9x9 dungeon size was too large for the uninitiated player.
Visual feedback is very important to the construction process. Doorways that were to be removed are highlighted in one colour, while another colour is used for those that are to be added. Tiles with any changes to their structure are marked in another different colour. When the prototype gets artwork from the Graphics team, these usability lessons will be kept firmly in mind so that the final experience is intuitive.
So, after a good round of iterative design and careful prototyping we have a dungeon builder we're proud of, and one that we think is great fun in its own right. It will be an integral part of Dungeon Assault II.
Mod Vile FunOrb Developer (Current Renown: Infinity+1)